Speech Emotion Recognition on RAVDESS (6-fold subject-independent CV)
93.4Weighted Accuracy (WA)Proposed Method
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Proposed MethodModality=Audio-only2025.12 | 93.4 | 92.28 | |
| Sun et al. [33]Modality=Multi-modal2025.12 | 87.99 | 87.96 | |
| Sadok et al. [36]Modality=Multi-modal2025.12 | 84.8 | — | |
| Yu et al. [34]Modality=Audio-only2025.12 | 81.86 | 82.75 | |
| Chumachenko et al. [35]Modality=Multi-modal2025.12 | 79.2 | — | |
| Chen et al. [13]Modality=Audio-only2025.12 | 75.36 | 75.28 | |
| Baevski et al. [32]Modality=Audio-only2025.12 | 74.38 | 73.44 | |
| Sun et al. [33]Modality=Audio-only2025.12 | 72.29 | 70.38 |